Calorimetric Study of Adsorption of Alkanes in High-Silica Zeolites

Previous studies of high-silica zeolites were mainly on MFI and FAU.1-12 ... ratio of 22 was manufactured by Philadelphia Quartz Corp. UTD-1 was suppl...
0 downloads 0 Views 114KB Size
J. Phys. Chem. B 1998, 102, 6865-6872

6865

Calorimetric Study of Adsorption of Alkanes in High-Silica Zeolites Scott Savitz, Flor Siperstein, Raymond J. Gorte, and Alan L. Myers Department of Chemical Engineering, UniVersity of PennsylVania, Philadelphia, PennsylVania 19104-6393 ReceiVed: April 13, 1998; In Final Form: June 23, 1998

Isosteric heats and adsorption isotherms were measured for combinations of three alkanes (methane, ethane, propane) on a series of six high-silica zeolites (TON, MTW, UTD-1, MFI, FER, FAU). Three of these zeolites (TON, MTW, UTD-1) have one-dimensional channels, two (MFI, FER) have two-dimensional intersecting channels, and FAU has spherical cavities. Since the adsorbate molecules are nonpolar and the zeolites possess a high Si/Al ratio, electrostatic energies are small in comparison to van der Waals energies. Heats of adsorption of methane at the limit of zero coverage are 27.2, 20.9, and 14.2 kJ/mol in TON, MTW, and UTD-1, respectively. This homologous series of zeolites has one-dimensional channels composed of 10-, 12-, and 14-membered rings of average effective diameter 5.0, 5.9, and 8.8 Å, respectively. Short-range gas-solid interactions between spherical methane molecules and the oxygen atoms composing the pore walls are explained by a smoothed integration of the Lennard-Jones 12-6 potential for cylindrical pores.

Introduction Nonbonding interactions between adsorbate molecules and zeolites consist of two principal types. The first type is longrange electrostatic interactions between adsorbate molecules and ions in the zeolite; the second type is short-range van der Waals (dispersion) interactions between the adsorbate molecules and the zeolitic framework. Electrostatic interactions are highly sensitive to the polarity of the adsorbate molecule and to the type and location of nonframework cations (e.g., Na, Ca, Mg); van der Waals interactions depend on the geometry of the microporous structure of the zeolite and the size and shape of the adsorbate molecules. Changes in the strength of electrostatic interactions are achieved by changing the silicon-to-aluminum ratio and by ion exchange of the nonframework cations of the zeolite. Dispersion interactions, which depend on the size and shape of the micropores, are less amenable to fine-tuning than the electrostatic interactions. Gas-solid dispersion interactions can be investigated by comparing zeolitic structures having the same geometry. The most straightforward example is the class of zeolites having one-dimensional channels of different size: TON, MTW, and UTD-1. Other structures such as ferrierite and silicalite have two-dimensional networks of intersecting channels of different size and shape; for these zeolites, dispersion interactions are somewhat more difficult to interpret. The presence of structural defects and residual aluminum may also complicate the interpretation of dispersion forces in high-silica materials. Previous studies of high-silica zeolites were mainly on MFI and FAU.1-12 Gases studied in these materials range from simple polar and nonpolar probes (Ar, CH4, Kr, Xe, N2, C2H4, and CO2) to long-chain alkanes and aromatics. Other zeolites that have been studied include TON, FER, KFI,13 and MOR.14 However, differences in the molecular structure of the probe molecules and the complexity of the pore structures in many of the molecular sieves examined make it difficult to isolate and quantify the effect of geometric factors upon dispersion forces. In the present study, work is focused on a homologous series of nonpolar probe molecules in a homologous series of

high-silica zeolites. In all of the systems examined, the dispersion energy dominates and electrostatic interactions are negligible or at least very small relative to dispersion. These results should be useful in establishing accurate force fields for dispersion interactions in computer simulations of adsorption. 2. Experimental Section Isosteric heats of adsorption and adsorption isotherms were measured for all combinations of three adsorbates (methane, ethane, and propane) and six adsorbents (TON, MTW, UTD-1, MFI, FER, FAU). The structure codes are listed in the Atlas of Zeolite Structure Types.15 All of these are siliceous zeolites with high Si/Al ratios and relatively few nonframework cations. Data for CH4 and C2H6 on silicalite published previously6 are included here for completeness. Materials. Ethane and propane with purities of 99% and 99.5%, respectively, were obtained from Air Products and Chemicals, Inc. Methane (99.99% pure) was obtained from Airco. The MFI (silicalite) powder used was Linde S115 manufactured by Union Carbide Corp. FER (ferrierite) with a Si/Al ratio of 22 was manufactured by Philadelphia Quartz Corp. UTD-1 was supplied by Ken Balkus and A. Ramsaran at the University of Texas, Dallas.16 FAU (siliceous faujasite) was prepared by Anthony Cheetham at the University of California, Santa Barbara.17 TON (ZSM-22) and MTW (ZSM-12) zeolites were synthesized in our laboratory according to published procedures.18-20 The Si/Al ratios and pore volumes for these zeolites are given in Table 1. Pore volumes were obtained from adsorption isotherms of n-hexane at 25 °C assuming that the density of adsorbed n-hexane at saturation is equal to its liquid density (0.655 g/cm3). The vapor pressure of n-hexane at 25 °C is 20.2 kPa, but saturation was reached at P < 3 kPa. Because of the assumption that liquid hexane completely fills the pores, the pore volumes extracted from the saturation capacity of n-hexane are only approximate. The powder samples were pressed into thin wafers to prevent accumulation of dust. Samples were weighed under ambient

S1089-5647(98)01836-7 CCC: $15.00 © 1998 American Chemical Society Published on Web 08/11/1998

6866 J. Phys. Chem. B, Vol. 102, No. 35, 1998

Savitz et al.

TABLE 1: Si/Al Ratios and Pore Volumes of Zeolites zeolite

Vp (cm3/g)

Si/Al ratio

TON MTW UTD-1 FER MFI FAU

0.069 0.097 0.106 0.105 0.173 0.245

52 140 ∞ 22 300 ∞

conditions, and the dehydrated weights were calculated from separate gravimetric measurements in vacuo. The degassing procedure for all zeolites was the same. Samples were heated under vacuum from room temperature to 110 °C over a period of 24 h for fresh samples and 12 h for used samples, then the temperature was ramped from 110 °C to 350 °C over a period of 12 h, and finally the temperature was held constant for 12 h. Apparatus. The combination calorimeter-volumetric apparatus contains a 20 cm3 cubical Pyrex sample cell encased on five of its six sides by thermopiles. The sample cell encased in thermopiles sits inside an aluminum block which provides a large thermal mass for rapid heat dissipation and small temperature rise. The output signal from the thermopiles, which is proportional to the heat flux, is input to a computer which stores the response curve and performs a numerical integration for the total heat released by adsorption of a known increment of gas. The primary calibration of the calorimeter is based upon the Clapeyron equation applied to a series of adsorption isotherms measured in a separate, high-precision volumetric apparatus for ethane on silicalite; the calibration based on electrical heating was essentially the same but less precise. The sample cell is connected to a dosing loop and pressure transducer for simultaneous determination of the adsorption isotherm and the heat of adsorption. The amount adsorbed is determined by the volumetric method as the difference between the amount of gas admitted and the amount of gas remaining in the dead space of the sample cell at equilibrium. Helium at room temperature was used to determine the dead space inside the sample cell for each zeolite sample. The reversibility of the measurements was verified by comparing points obtained by adsorption and desorption of the sample. The experimental results were checked for reproducibility by repeating the experiment for different samples of the same adsorbent. 3. Results of Experiment The series of zeolites TON, MTW, and UTD-1 is especially interesting because all three have one-dimensional straight channels of increasing ring size, as shown in Table 2. Isosteric heats of adsorption for this series are shown on Figures 1-3 for CH4, C2H6, and C3H8, respectively. The most prominent feature of these results is the decrease in heat of adsorption with increasing pore size. However, a more subtle competition between energetic heterogeneity of gas-solid interactions and cooperative gas-gas interactions is evident in the heat profiles. Energetic heterogeneity generates a heat of adsorption which decreases with loading; cooperative interactions generate a heat of adsorption which increases with loading. As expected, the relative strength of cooperative interactions increases in the order CH4, C2H6, C3H8 for a given adsorbent and in the order of increasing pore size for a given adsorbate. Isosteric heats of adsorption are reported for FER, MFI, and FAU in Figures 4-6, respectively. The trend of lower heats

TABLE 2: Cavity Dimensions of Zeolitesa

zeolite

type of cavity

TON (ZSM-22) MTW (ZSM-12) UTD-1 FER MFI FAU

1-D cylindrical 1-D cylindrical 1-D cylindrical 2-D cylindrical 2-D cylindrical spherical

no. of members in ring 10 12 14 8,10 10,10

cavity 5.5 × 4.4 6.2 × 5.5 10.0 × 7.5 4.8 × 3.5, 5.4 × 4.2 5.3 × 5.6, 5.1 × 5.5 12.6

a The dimensions are the major and minor axes of the elliptical channels. For FER and MFI, dimensions are given for both channels. For FAU, the dimension is the diameter of the supercage. In all cases, the distance is the center-to-center distance between oxygen atoms minus 2.7 Å.

Figure 1. Isosteric heats of adsorption of methane on siliceous materials with one-dimensional channels at 298 K: (4) TON; (0) MTW; (3) UTD-1.

in larger cavities is apparent in this series of zeolites. Bates et al.21 showed that for pore openings larger than 5 Å, heats of adsorption decrease as the opening of the pore increases. The subtler competition between energetic heterogeneity and cooperative interactions mentioned previously causes the relative strength of cooperative interactions to increase with pore size. Isosteric heats are reported in Tables 9-26. Limiting values at zero coverage for the heats of adsorption are listed in Table 3. This is the Boltzman average energy (strictly enthalpy) of a single probe molecule interacting with the walls of the channel. As shown in Table 4, our heats agree very well with the data of Zhu et al.9 and the C3H8/TON results of Eder and Lercher.13 All of the published data agree within 5% for CH4, C2H6, and C3H8 on MFI. However, there is a variation of (10% in heats of adsorption reported by different investigators for the other systems. The one-dimensional straight-channel zeolites TON, MTW, and UTD-1 exhibit a strong decrease in the zero-coverage heat of adsorption with increasing pore size, as expected for dispersion energies. Similarly, the two-dimensional intersecting channels in FER and MFI show a strong decrease in heat of adsorption with increasing pore size. The spherical shape of the supercavities in FAU makes a comparison with the cylindrical channels of the other zeolites problematic, but the heat of adsorption for all the probe molecules is least in FAU.

Adsorption of Alkanes in High-Silica Zeolites

Figure 2. Isosteric heats of adsorption of ethane on siliceous materials with one-dimensional channels at 298 K: (4) TON; (0) MTW; (3) UTD-1.

Figure 3. Isosteric heats of adsorption of propane on siliceous materials with one-dimensional channels at 298 K: (4) TON; (0) MTW; (3) UTD-1.

It is interesting that the increase in heat of adsorption per additional carbon atom in the alkane chain is a strong function of the size of the micropore. It has been reported from experimental measurements7,8,11,13 and from computer simulations21-26 that heats of adsorption in MFI increase approximately 10 kJ/mol with each carbon atom in the alkane chain. Table 5 confirms these results: the incremental increase in heat of adsorption per carbon atom is about 10 kJ/mol for MFI but increases in smaller pores (FER) and decreases in larger pores (UTD-1). Henry’s constants calculated from adsorption isotherms measured in a volumetric apparatus are given in Table 6. The adsorption isotherms are reported in Tables 27-37; asterisks identify desorption points taken to check for reversibility.

J. Phys. Chem. B, Vol. 102, No. 35, 1998 6867

Figure 4. Isosteric heats of adsorption of alkanes on FER at 298 K: (n) methane; (4) ethane; (0) propane.

Figure 5. Isosteric heats of adsorption of alkanes on MFI at 298 K: (n) methane; (4) ethane; (0) propane.

4. Discussion and Interpretation of Results The heats of adsorption obtained for zeolites in onedimensional channels (TON, MTW, UTD-1) were compared with theoretical calculations based upon the following model: (1) Lennard-Jones 12-6 potential (parameters 12 and σ12) for the energy of an adsorbate molecule interacting with a single oxygen atom in the wall of a zeolite. (2) Integration of the potential energy over a single layer of oxygen atoms distributed uniformly in the wall of a cylindrical, one-dimensional pore. The potential energy for this model27 is

E(r) )

{ ( ) [ ( ) ] ( ) [ ( ) ]}

5 21 σ12 π1s 2 32 R

10



r

∑Rk R k)0

2k

-

σ12 R

4



r

∑ βk R k)0

2k

(1)

6868 J. Phys. Chem. B, Vol. 102, No. 35, 1998

Savitz et al. TABLE 5: Incremental Increase in Isosteric Heat Per Methyl or Methylene Group (∆Qst) Compared to Average Pore Diameter (D) of High-silica Zeolites zeolite

D (Å)

∆Qst (kJ/mol)

FER TON MFI MTW UTD-1 FAU

4.5 5.0 5.4 5.9 8.8 12.6

12.8 10.8 10.2 8.4 7.0 7.6

TABLE 6: Henry’s Constants (H) of Alkanes on Siliceous Zeolites methane H mmol/ (g-torr)

T (°C)

H mmol/ (g-torr)

T (°C)

FER TON MFI MTW UTD-1

0.0026 0.00091 0.00084a 0.00054 0.00042

35.8 35.9 23.0 36.8 36.0

0.096 0.0188 0.0164a 0.0124 0.0021

35.8 35.9 23.0 37.0 36.1

CH4

C2H6

C3H8

27.7 27.2 20.9 20.9 14.2 11.0

41.7 39.0 31.1 29.5 22.2 19.7

53.3 48.8 41.4 37.6 28.1 26.2

Qst (kJ/mol) gas

a

zeolite

this work

other

MFI

20.9

C2H6

MFI

31.1

C3H8

MFI

41.4

C2H6

FAU

19.7

C3H8

FAU

26.2

C3H8 C3H8

FER TON

53.3 48.8

20.5 [8] 20 [5] 19 [9] 33 [8] 30 [5] 32 [1] 31 [9] 41 [8] 41 [5] 41 [9] 24 [12] 22 [10] 30 [12] 28 [10] 49 [13] 49 [13]

3.67

36.2

0.441

36.3

0.0150

36.3

6 1s ) πF12σ122 5

(4)

where F is the number of oxygen atoms per unit area in the wall of the cylinder. 1s is the well depth at the potential-energy minimum in the limit R f ∞, i.e., the well depth adjacent to a plane surface. The collision diameter corresponding to U ) 0 in the Lennard-Jones 12-6 potential is σ12 ) (2.700 + 3.691)/2 ) 3.195 Å, based upon 2.700 Å for the O atom15 and 3.691 Å for the CH4 molecule.28 The value of the well-depth 12/k ) 133.33 K was reported previously.29 Since the actual channels are elliptical, the average of the major and minor axes (plus 2.7 Å for the diameter of an O atom) was taken as the diameter (D) of the cylindrical channel in the theory so that R ) D/2 ) 3.825, 4.275, and 5.725 for TON, MTW, and UTD-1, respectively. The area density of oxygen atoms in Table 7 was obtained by counting the oxygen atoms in the channel of a unit cell and dividing by the area of the elliptical channel (2πL[(R12 + R22)/2]1/2). From eq 4, 1s/k ) 854, 917, and 946 K for TON, MTW, and UTD-1, respectively. Figure 7 shows the potential energies calculated from eq 1 for the three zeolites with one-dimensional channels using the values for F, R, σ12, and 12 given above. The potential minima -U/k ) 2540, 2150, and 1540 K for TON, MTW, and UTD-1, respectively. The location of the minimum is at the center of the channel for TON but moves progressively toward the wall for the wider channels in MTW and UTD-1. The adsorption second virial coefficient for a cylindrical channel is

TABLE 4: Comparison of Isosteric Heats of Adsorption with Previous Worka

CH4

T (°C)

the axis. The gas-solid energy parameter 1s is a function of the surface density of oxygen atoms

Qst (kJ/mol) zeolite

H mmol/ (g-torr)

Data from ref 6.

TABLE 3: Zero-Coverage Isosteric Heats of Adsorption of Probe Molecules in High-Silica Zeolites at 25 °C

FER TON MFI MTW UTD-1 FAU

propane

zeolites

a

Figure 6. Isosteric heats of adsorption of alkanes on FAU at 298 K: (n) methane; (4) ethane; (0) propane.

ethane

The number given in brackets is the reference number.

where

B(T) )

Γ(-4.5) ) Γ(-4.5 - k)Γ(k + 1)

(2)

Γ(-1.5) βk1/2 ) Γ(-1.5 - k)Γ(k + 1)

(3)

Rk

1/2

R is the radius of the cylinder defined by the centers of the oxygen atoms, and r is the distance of the gas molecule from

∫0R(e-E(r)/kT - 1)r dr

2πL M

(5)

where L is channel length and M is the mass of a unit cell15,16 given in Table 7. The -1 in the integrand is required to correct absolute adsorption to excess adsorption and corresponds to the experimental measurement of surface excess (the actual amount adsorbed less the amount present when the surface does not adsorb).

Adsorption of Alkanes in High-Silica Zeolites

J. Phys. Chem. B, Vol. 102, No. 35, 1998 6869

TABLE 7: Structural Properties of Zeolites with One-Dimensional Elliptical Channels zeolite

g/mol (u.c.)

O atoms/ channel

channels/ u.c.

L (Å)

R1 (Å)

R2 (Å)

R (Å)

Area (Å2)

F (O atoms/Å2)

TON MTW UTD-1

1442 1683 3846

20 24 56

2 1 2

5.0 5.0 8.4

3.55 4.10 5.10

4.10 4.45 6.35

3.825 4.275 5.725

120.5 134.4 304.0

0.166 0.179 0.184

TABLE 10: Methane on MTW n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.036 0.088

20.54 19.97

0.151 0.222

19.06 18.49

TABLE 11: Methane on UTD-1 n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.021 0.053

13.63 13.94

0.077 0.113

13.69 12.39

TABLE 12: Methane on MFI

Figure 7. Potential energy of a methane molecule in high-silica zeolites as a function of the dimensionless radius (r/R) of a cylindrical pore: (‚‚‚) TON; (- - -) MTW; (s) UTD-1.

TABLE 8: Comparison of Theory with Experiment for Henry’s Constant and Heat of Adsorption of Methane in High-Silica Zeolites with One-Dimensional Pores qst (kJ/mol)

H (mmol/g-Torr)

zeolites

theory

exp

theory

exp

TON MTW UTD-1

22.4 19.7 14.2

27.2 20.9 14.2

0.00106 0.00040 0.00023

0.00091 0.00054 0.00042

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.019 0.084 0.157

26.91 25.75 24.65

0.234 0.308 0.380

24.20 24.35 23.31

0.437 0.501

21.60 19.66

The energy of adsorption at the limit of zero coverage is

E h)

)r dr ∫0 E(r)(e ∫0R(e-E(r)/kT - 1)r dr

(6)

The zero-coverage isosteric heat (difference of enthalpies) is

(7)

Henry’s constants are calculated from the second virial coefficient:

H)

B kT

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.007 0.025 0.048 0.076 0.103 0.156 0.161 0.182 0.187 0.212 0.238

21.46 21.36 20.66 20.84 21.01 20.75 21.67 20.88 20.98 21.14 20.98

0.264 0.288 0.312 0.336 0.359 0.382 0.426 0.447 0.468 0.489 0.510

20.64 20.78 20.92 21.02 21.22 21.08 20.66 21.06 21.39 21.30 21.36

0.530 0.550 0.550 0.569 0.570 0.589 0.608 0.626 0.628 0.645

21.25 21.16 21.09 21.06 21.88 21.62 21.35 21.58 21.87 20.48

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.059 0.191 0.311

27.82 27.42 27.11

0.441 0.571 0.704

26.97 26.50 27.22

0.831

26.44

TABLE 14: Methane on FAU n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.020 0.071

10.85 10.39

0.129 0.193

9.94 9.70

0.268 0.344

8.92 7.25

TABLE 15: Ethane on TON

-E(r)/kT

h + kT qst ) -E

Qst (kJ/mol)

TABLE 13: Methane on FER

TABLE 9: Methane on TON

R

n (mol/kg)

(8)

Zero-coverage isosteric heats and Henry’s constants calculated from eqs 1-8 are summarized in Table 8. The average absolute error is 8% for the heats and 29% for the Henry’s constants. This error, although larger than the experimental uncertainty,

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.072 0.168 0.274 0.394

39.43 37.41 36.89 35.95

0.495 0.609 0.707 0.797

36.06 34.74 35.80 34.95

0.877 0.935 0.986 1.036

34.57 32.79 32.93 32.00

is surprisingly small considering that published values of Lennard-Jones interaction constants for CH4-O were used. The 29% error between experimental and calculated Henry’s constants is actually in reasonable agreement considering that the Henry’s constant is an exponential function of the gas-solid adsorption energy. For example, for CH4 adsorbed in MTW, an error of only 5% in the potential-well depth corresponds to an error of 45% in the Henry constant. One would expect the neglect of oxygen atoms not located in the walls of the channel to require an artificially large value of 12. Perhaps the neglect of long-range dispersion interactions is compensated by other approximations such as smooth walls and circular instead

6870 J. Phys. Chem. B, Vol. 102, No. 35, 1998

Savitz et al.

TABLE 16: Ethane on MTW

TABLE 23: Propane on UTD-1

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.084 0.181 0.272

28.98 29.60 30.73

0.384 0.490 0.595

29.61 29.04 29.81

0.681 0.733 0.813

28.88 30.11 29.97

0.104 0.214 0.321 0.422 0.518

28.74 29.83 30.07 30.71 31.02

0.609 0.656 0.730 0.813 0.889

31.41 32.11 31.89 32.52 32.87

0.954 1.024 1.085 1.147 1.205

33.43 34.00 33.95 35.09 36.22

TABLE 17: Ethane on UTD-1 n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.017 0.097 0.146 0.293

23.08 21.04 22.05 24.08

0.375 0.458 0.539 0.623

21.38 22.97 22.84 22.47

0.709 0.791 0.808 0.892

23.00 22.57 23.79 22.95

TABLE 18: Ethane on MFI n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.156 0.213 0.298 0.382 0.468 0.574 0.686 0.780 0.800

31.32 30.13 31.51 31.69 31.29 31.51 31.40 31.98 31.46

0.892 1.000 1.102 1.203 1.297 1.385 1.465 1.541 1.602

31.50 31.67 31.84 32.10 32.00 32.19 32.04 32.14 33.33

1.607 1.662 1.666 1.715 1.761 1.802 1.837

32.17 32.04 32.25 32.52 31.52 31.49 31.06

TABLE 24: Propane on MFI n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.080 0.168 0.258 0.346 0.435 0.525

41.59 42.10 41.68 42.16 42.74 43.17

0.616 0.701 0.793 0.887 0.992 1.111

41.99 42.75 43.33 43.82 45.07 44.63

1.240 1.359 1.470 1.573 1.674

44.96 45.19 45.40 45.65 44.91

TABLE 25: Propane on FER n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.097 0.207 0.324 0.454

52.77 51.60 51.32 50.77

0.592 0.734 0.880 1.019

50.48 48.98 48.32 47.29

1.157 1.274 1.382

45.39 46.35 45.13

TABLE 26: Propane on FAU TABLE 19: Ethane on FER n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.122 0.238 0.352 0.472 0.596

41.41 41.36 40.81 40.15 40.13

0.721 0.864 0.996 1.127 1.260

39.92 39.41 39.12 38.81 38.47

1.389 1.497 1.593

38.19 37.83 37.55

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.100 0.263 0.492 0.792 1.027

27.34 26.68 26.14 27.00 27.09

1.267 1.516 1.759 1.997 2.204

27.92 28.55 29.74 29.33 30.15

2.408 2.591 2.768 2.974 3.158

30.05 30.25 30.42 31.47 29.99

TABLE 27: Methane on TON at 35.9 °C

TABLE 20: Ethane on FAU n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.109 0.227 0.342 0.488

19.60 19.99 18.83 19.47

0.663 0.884 1.085 1.324

19.70 19.82 19.83 20.18

1.593 1.855 2.055

20.45 19.70 18.41

TABLE 21: Propane on TON n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.072 0.160 0.250 0.353

49.20 48.09 47.05 46.37

0.462 0.570 0.659 0.738

46.75 47.30 47.06 46.55

0.803 0.864 0.967 1.006

45.49 45.92 44.72 45.25

TABLE 22: Propane on MTW n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

n (mol/kg)

Qst (kJ/mol)

0.124 0.206

37.32 36.39

0.255 0.340

37.67 37.21

0.408 0.433

37.55 35.49

of elliptical channels. It has been shown30 that the detailed structure of the micropore becomes important as the pore size decreases. Figure 8 shows a comparison of theory with the experimental data for all six high-silica zeolites. It is emphasized that we have not fitted any parameters to the experimental data; values of all parameters were obtained either from the structure of the zeolites15 or from previously published potential parameters.29 The solid line is the isosteric heat of methane calculated from

P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

12.63 21.49 31.54 36.58 43.71 59.25 73.95 89.24

0.0110 0.0191 0.0279 0.0326* 0.0384 0.0508 0.0616 0.0731

108.4 133.0 152.2 187.3 235.1 279.0 279.7 337.5

0.0867 0.1033 0.1160 0.1381 0.1653 0.1893* 0.1898 0.2189

338.5 390.3 407.2 435.2 474.3 546.1 632.7 703.4

0.2198 0.2455 0.2549 0.2656 0.2828 0.3144 0.3461 0.3722

P (torr)

n (mmol/g)

320.0 450.3 599.3 829.9

0.1805 0.2322 0.3084 0.3856

TABLE 28: Methane on MTW at 36.8 °C P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

7.32 12.53 20.24 26.92 34.74 35.08

0.0038 0.0070* 0.0107 0.0147 0.0196* 0.0192

44.01 73.02 96.36 96.78 141.80 214.40

0.0240 0.0386 0.0505 0.0556 0.0783 0.1131

eqs 6 and 7 as a function of the radius (R) of the pore for an average value 1s/k ) 907 K. For the experimental points, the radius of the pore is defined as one-half of the average of the major and minor axes of one-dimensional elliptical pores. For FER and MFI, the radius is defined as the average radius of both elliptical pores; for FAU, the radius is defined as the radius of the supercavity. To conform with the model, these pore diameters are between oxygen-atom centers; the effective diameters in Table 1 are 2.7 Å less to account for the size of the oxygen atoms. The agreement of theory with experiment is much better than expected, considering the simplicity of the

Adsorption of Alkanes in High-Silica Zeolites

J. Phys. Chem. B, Vol. 102, No. 35, 1998 6871

TABLE 29: Methane on UTD-1 at 36.0 °C P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

57.98 63.77 80.59

0.0211 0.0223 0.0266*

92.50 95.19 134.20

0.0309 0.0302 0.0399

TABLE 34: Ethane on FER at 35.8 °C

P (torr)

n (mmol/g)

161.2

0.0471

P (torr)

n (mmol/g)

393.6 464.3 497.2 547.5 563.3 688.9 799.5

0.6097 0.6807 0.7111* 0.7553 0.7676 0.8640 0.9379

TABLE 30: Methane on FER at 36.0 °C P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

5.27 13.65 16.21 22.19 29.12 33.18 41.02 54.02 72.91

0.0135 0.0323 0.0405* 0.0508 0.0678 0.0759 0.0924 0.1183 0.1546

88.29 90.40 110.80 111.50 141.30 172.80 219.90 271.30 327.90

0.1836* 0.1866 0.2234 0.2244 0.2744 0.3337 0.4014 0.4693 0.5383

P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

1.99 5.00 8.30 9.70 10.47 12.58 13.43 14.59 16.72 18.30

0.0362 0.0812 0.1223 0.1385* 0.1484* 0.1678 0.1766* 0.1875 0.2080 0.2206

23.91 32.68 41.95 52.19 63.89 76.96 78.83 89.68 95.15 112.00

0.2656 0.3250 0.3780 0.4290 0.4779 0.5250 0.5322* 0.5645 0.5803 0.6225

116.7 136.8 159.2 190.8 301.0 362.8 408.9 479.1 569.6 676.6

0.6337 0.6747 0.7133 0.7593 0.8655 0.9059 0.9309 0.9624 0.9946 1.0246

TABLE 32: Ethane on MTW at 37.0 °C n (mmol/g)

P (torr)

n (mmol/g)

3.48 8.42 15.39 22.77

0.0387 0.0899 0.1557 0.2185

23.94 32.45 35.79 50.38

0.2287* 0.2928 0.3160 0.4078

P (torr)

n (mmol/g)

56.66 70.92

0.4423* 0.5137

P (torr)

n (mmol/g)

354.4 355.8 482.9 490.8 580.0 693.8

0.6446 0.6489 0.8423 0.8457 0.9692 1.1260

TABLE 33: Ethane on UTD-1 at 36.1 °C P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

33.45 47.87 51.99 63.17 66.36 82.60 86.45

0.0645 0.0929* 0.1007 0.1224 0.1287 0.1595 0.1691

118.5 141.3 163.8 193.6 218.9 264.4 265.9

0.2327 0.2711 0.3185 0.3674 0.4162 0.4941 0.4973*

n (mmol/g)

P (torr)

n (mmol/g)

0.50 1.51 3.39 6.21 10.61 15.36 17.61 21.03

0.0438 0.1307 0.2470 0.3862 0.5502 0.6840* 0.7345 0.8058

25.90 27.46 28.90 31.81 34.75 36.27 37.55 40.97

0.8978* 0.9121 0.9372* 0.9808* 1.0163 1.0258 1.0435 1.0805*

P (torr)

n (mmol/g)

46.68 47.00 56.25 60.51 73.99 88.29

1.1343 1.1347 1.2058 1.2361 1.3123 1.3779

TABLE 35: Propane on UTD-1 at 36.3 °C P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

7.21 15.11

0.1053 0.2225

18.05 23.39

0.2700* 0.3468

27.10 39.43

0.4046 0.5713

P (torr)

n (mmol/g)

10.04 10.49 13.86 16.01 27.50 57.24 224.50 472.50 756.20

2.0292 2.0691 2.3353 2.4558 2.8704 3.2460 3.7043 3.8584 3.9584

P (torr)

n (mmol/g)

95.45 251.40 337.90 443.90 577.80 858.10

5.9847 6.4155 6.5045* 6.5853 6.6577 6.7542

TABLE 36: Propane on MFI at 36.3 °C

TABLE 31: Ethane on TON at 35.9 °C

P (torr)

P (torr)

P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

0.19 0.37 0.60 0.72 1.03 1.03 1.20 1.62 1.65 1.81

0.0801 0.1522 0.2368 0.2760* 0.3825 0.3261* 0.4477 0.5233 0.5007* 0.6437

1.96 2.45 2.54 2.73 3.81 3.94 5.93 6.48 7.15 7.52

0.6090* 0.7668 0.7886 0.8527 1.0999* 1.1201 1.5016 1.5841* 1.6948 1.7245

TABLE 37: Propane on FER at 36.2 °C P (torr)

n (mmol/g)

P (torr)

n (mmol/g)

0.03 0.11 0.23 0.25 0.36 0.41 0.54 0.79

0.0769 0.3362 0.6391 0.6768 0.8989* 0.9890 1.1958 1.5373

1.23 1.41 1.82 2.31 3.97 6.13 15.48 46.24

1.8791 2.0354* 2.3390 2.6066 3.2678 3.7813 4.7580 5.6104

model for short-range gas-solid interactions. A detailed atomatom model should improve the agreement with experiment, especially for the small channels in FER and TON. 5. Conclusions We studied the adsorption of a homologous series of alkanes (C1-C3) on a homologous series of high-silica zeolites with straight one-dimensional channels (TON, MTW, UTD-1). The zero-coverage heats of adsorption are inversely proportional to the pore diameter and agree approximately with a smoothed integration for short-range van der Waals interactions. Heats of adsorption of high-silica zeolites with two-dimensional intersecting channels (FER, MFI) and spherical supercages (FAU) also agree with the model. The incremental heat of adsorption per alkane carbon atom is about 10 kJ/mol for MFI but varies between 8 and 13 kJ/mol depending on the pore size.

Figure 8. Comparison of theory with experiment for heat of adsorption of methane in high-silica zeolites at 298 K. Solid line calculated from eqs 1-4, 6, and 7 with 1s/k ) 907 K and σ ) 3.195 Å. Experimental points are from Table 3.

Acknowledgment. Support was provided by NSF CTS-9610030. We also thank D.H. Olson, who supplied the FER and TON samples, and Ken Balkus, who supplied the UTD-1 sample.

6872 J. Phys. Chem. B, Vol. 102, No. 35, 1998 References and Notes (1) Thamm, H. Zeolites 1987, 7, 341. (2) Abdul-Rehman, H. B.; Hasanain, M. A.; Loughlin, K. F. Ind. Eng. Chem. Res. 1990, 29, 1525. (3) Golden, T. C.; Sircar, S. J. Colloid Interface Sci. 1994, 162, 182. (4) Otto, K.; Montreuil, C. N.; Todor, O.; McCabe, R. W.; Gandhi, H. S. Ind. Eng. Chem. Res. 1991, 30 (10), 2333. (5) Hufton J. R.; Danner, R. P. AIChE J. 1993, 39 (6), 954. (6) Dunne, J. A.; Mariwala, R.; Rao, M.; Sircar, S.; Gorte, R. J.; Myers, A. L. Langmuir 1996, 12, 5888. (7) Sun, M. S.; Talu, O.; Shah, D. B. J. Phys. Chem. 1996, 100, 17276. (8) Sun, M. S.; Shah, D. B.; Xu, H. H.; Talu, O. J. Phys. Chem. B 1998, 102, 1466. (9) Zhu, W.; van de Graaf, J. M.; van den Broeke, L. J. P.; Kapteijn, F.; Moulijn, J. A. Ind. Eng. Chem. Res. 1998, 37 (5) 1934. (10) Stach, H.; Lohse, U.; Thamm, H.; Schirmer, W. Zeolites 1986, 6, 74. (11) Stach, H.; Fiedler, K.; Ja¨nchen J. Pure Appl. Chem. 1993, 65 (10), 2193. (12) Hampson, J. A.; Rees, L. V. C. J. Chem. Soc., Faraday Trans. 1993, 89 (16), 3169. (13) Eder, F.; Lercher, J. A. J. Phys. Chem. B 1997, 101, 1273. (14) Triebe, R. W.; Tezel, F. H.; Khulbe, K. C. Gas Sep. Purif. 1996, 10 (1), 81. (15) Olson, D. H.; Meier, W. M. Atlas of Zeolite Structure Types, 3rd ed.; Butterworth-Heinemann: London, 1992.

Savitz et al. (16) Lobo, R. F.; Tsapatsis, M.; Freyhardt, C. C.; Khodabandeh, S.; Wagner, P.; Chen, C. Y.; Balkus, K. J.; Zones, S. I.; Davis, M. E. J. Am. Chem. Soc. 1997, 119, 8474. (17) Milton, R. M. U.S. Patent 2,882,244, 1959. (18) Kofke, T. J. G.; Gorte, R. J.; Kokotailo, G. T.; Farneth, W. E. J. Catal. 1989, 115, 265. (19) Rosinski, E. J.; Rubin, M. K. U.S. Patent 3,832,449, 1974. Givens, E. N.; Plank, C. J.; Rosinski, E. J. U.S. Patent 4,052,472, 1977. (20) Valyocsik, E. W. U.S. Patent 4,481,177, 1984. (21) Bates, S. P.; van Well, W. J. M.; van Santen, R. A.; Smit, B. J. Am. Chem. Soc. 1996, 118, 6753. (22) Maginn, E. J.; Bell, A. T.; Theodorou, D. N. J. Phys. Chem. 1995, 99, 2057. (23) Smit, B.; Siepmann, J. I. J. Phys. Chem. 1994, 98, 8442. (24) Bates, S. P.; van Well, W. J. M.; van Santen, R. A.; Smit, B. J. Phys. Chem. 1996, 100, 17573. (25) Smit, B. J. Phys. Chem. 1995, 99, 5597. (26) June, R. L.; Bell, A. T.; Theodorou, D. N. J. Phys. Chem. 1990, 94, 1508. (27) Everett, D. H.; Powl, J. C. J. Chem. Soc., Faraday Trans. 1 1976, 72, 619. (28) Prausnitz, J. M.; Myers, A. L. AIChE J. 1963, 9 (1), 5. (29) Goodbody, S. J.; Watanabe, K.; MacGowan, D.; Walton, J. P. R. B.; Quirke, N. J. Chem. Soc., Faraday Trans. 1991, 87, 1951. (30) Curry, J. E.; Cushman, J. H. Mol. Phys. 1995, 85 (1), 173.